1. Field of the Invention
The present invention relates to electroless deposition of copper (or possibly an alloy predominating in copper) from a solution in which copper ions are dissolved, in order to provide a metal deposit or film on a desired, suitably-prepared, substrate when immersed in or contacted by the solution, without the employment of external electrical energy to bring about such reduction. The invention relates more particularly to electroless copper baths employing a non-formaldehyde type reducing agent, and more particularly a soluble hypophosphite reducing agent, for effecting conversion of the copper ions to copper metal in order to form adherent, highly conductive metal films on controlled surfaces of substrates, particularly nonconductive substrates.
2. Description of the Prior Art
The conventional electroless plating art as commercially practiced in the deposition of copper onto various substrates, especially nonconductive substrates, almost without exception today uses highly alkaline formaldehyde solutions of divalent copper complexed with various well-known agents such as Rochelle salt, amines and others. A current survey of the practical art is summarized in an article entitled "Electroless Copper Plating", by Purhpavanam and Shenoi, published in "Finishing Industries", October 1977, pages 36 et seq. The article lists the various components of electroless copper plating solutions, and discusses useful alternatives in each category. With respect to available agents for reducing the copper ion of the bath, the article lists hypophosphites, phosphites, hyposulfites, sulfites, sulfoxylates, thiosulfates, hydrazine, hydrazoic acid, axides, formaldehyde, formate and tartrate as having been tried. Hypophosphite is stated to be "very effective in alkaline or acid solutions", but the article does not define what is meant by this and goes on immediately to report that "this operates only at higher temperatures and under these conditions there appears to be a rapid reduction of copper in the bulk of solution." In other words decomposition of the solution occurs, resulting in the bath being of no further use for electroless plating. Other reducing agents from the abovementioned list are also discussed, more particularly hydrazine, borohydride and dimethylamine borane. The article states that "The best reducing agent for copper is considered to be formaldehyde." and later concludes that "No other reducing agent is capable of replacing formaldehyde and hence on(sic) (only?) the Fehlings-formaldehyde solution with modifications is maintaining its superior position in electroless copper plating."
In an article entitled "Fabrication of Semitransparent Masks", Feldstein and Weiner, J. Electrochemical Soc., Vol. 120, pp 1654-1657 (December 1973), the use of hypophosphite reducing agent is described in connection with production of semitransparent resists or masks, using an alkaline copper sulfate, EDTA-complexed, bath. The article indicates that the resulting film deposited on a catalyzed substrate immersed in such bath is cuprous oxide (Cu.sub.2 O), and concludes that reduction of copper ions to metallic copper does not take place to any appreciable extent in a hypophosphite-reduced system. The article further reports that the deposited cuprous oxide does not provide sufficient catalytic activity for continuation of the plating process.
An earlier study entitled "Electroless Copper Plating in Printed Circuitry", E. B. Saubestre, The Sylvania Technologist, Vol. XII, No. 1, January 1959, also considered the reactions of copper ions in solutions containing a hypophosphite reducing agent, and reported work on attempted reduction of copper in alkaline hypophosphite solution as well as in alkaline hyposulfite and formaldehyde solutions. In order to obtain copper by chemical reduction, it was found necessary by the author to have either a system in which there is little tendency for the cuprous ion to form, or one in which the cuprous ion is rendered soluble by formation of a suitable complex ion. Of the various solutions tested, only the following four combinations were found to offer promise:
(a) Fehling's solution with formaldehyde PA1 (b) Fehling's solution with hydrazine sulfate PA1 (c) Acid sulfate solution with sodium hypophosphite PA1 (d) Acid sulfate solution with sodium hyposulfite.
It was reported that investigation of these possibilities revealed that copper is a pronounced reduction catalyst only in the Fehling's-formaldehyde solution, so further work was accordingly concentrated along that line. Supplementing this article is another by the same author which appears in Technical Proceedings of the Golden Jubilee Convention of American Electroplaters Society, Vol. 46, pages 264 et seq; 1959. In this article a comprehensive review is presented on reducing agents for copper, and particularly sodium hypophosphite in a series of different types of copper solutions. The conclusion reached was that "In general, this reducing agent shows little promise except in Fehling's and sulfate solutions operated at high temperatures and high hypophosphite concentrations. However, under these conditions, there appears to be rapid reduction of copper in the bulk of the solution as well." In other words the solutions decompose and cannot be used on a continuing basis and particularly not over an extended period of time. Hyposulfite was also investigated and the conclusion reached was that it "is more effective than hypophosphite, but again, since deposition tends to occur throughout the solution, this reducing agent probably lends itself only to spraying applications". That is, one involving continuous spraying of separate streams, one containing copper ions, the other the reducer. Such conditions of operation are commercially noneconomic and totally impractical.
The technical literature clearly establishes that while hypophosphite agents are effective and universally used as reducing agents in electroless nickel deposition techniques, they have not been found useful practically for electroless copper deposition. For copper, formaldehyde is the overwhelming choice in commercial plating today. The only viable alternatives even mentioned are borohydride, dimethylamine borane and hydrazine.
The patent literature confirms the foregoing practical experience and conclusion. U.S. patents directed specifically to electroless copper issuing between 1960 and 1977 almost invariably list formaldehyde or formaldehyde precursors, many times giving these as the only reducing agents although borohydrides and boranes appear in several patents, and there is occasional reference to hydrazine. There are a few references to alkali metal hypophosphites and hydrosulfites; but in the case of hypophosphites the disclosures relate solely to acid solutions operating at pH levels of 3.0 or less. For example, U.S. Pat. No. 3,046,159 mentions the use of hypophosphite reducing agents in plating by chemical reduction from a solution containing a normally insoluble copper compound, such as cupric oxide, in conjunction with an ammoniacal compound such as ammonium sulfate or ammonium chloride, to which sodium hypophosphite is added as the reducing agent. In all examples the solution is strongly acid (pH 3.0 or less). In order to increase the plating rate the patent recommends that the solution temperature be increased, but also recognizes that this leads to instability and great difficulty in preventing complete collapse of the system. Attempts to duplicate the teaching of this patent using standard, properly cleaned copper-clad panels, have produced only a brownish oxide deposit. When the teaching is applied to a nonmetallic substrate, such as a standard ABS of platable grade suitably prepared (catalyzed) for electroless plating, the cupric oxide particles in the bath form on the surface along with a reddish, non-adherent deposit which rubs off on the fingers when touched. Attempts to electroplate the coated substrate failed completely because the deposit simply burns off, proving that it is essentially non-conductive, leading to the conclusion that it is not metallic copper or at least is not significantly so.
It is interesting to note that other U.S. Pat. Nos., such as 3,403,035; 3,443,988; 3,485,653; 3,515,563; 3,615,737; and 3,738,849, these being the only others currently known to the present inventors which contain reference to hypophosphites as reducing agents in electroless copper bath, also relate to strongly acid copper solutions. It is clear from these patent disclosures that alkaline formaldehyde systems, which are generally always also mentioned, are those actually considered to be useful in practice.
A recent U.S. Pat. No. 4,036,651 teaches incorporation of sodium hypophosphite as a "plating rate adjuster" in an alkaline formaldehyde type electroless copper solution. The patent states expressly "Although sodium hypophosphite is, itself, a reducing agent in electroless nickel, cobalt, palladium and silver plating baths, it is not a satisfactory reducing agent (i.e., will not reduce Cu.sup.++ .fwdarw.Cu.degree.) when used alone in alkaline electroless copper plating baths. In the baths of the present invention [U.S. Pat. No. 4,036,651], the sodium hypophosphite is not used up in the plating reaction. Instead, it appears to act as a catalyst." (Bracketed insert added).
In the prior patents, where both electroless nickel as well as copper baths are disclosed, the bath composition examples invariably employ formaldehyde-type reducing agents for the copper formulations and, in contrast, hypophosphites for the nickel formulations. There is no suggestion in the patent art that the hypophosphite of the nickel baths could be substituted for formaldehyde in copper baths. See U.S. Pat. Nos. 3,370,974; 3,379,556; 3,617,363; 3,619,243; 3,649,308; 3,666,527; 3,668,082; 3,672,925; 3,672,937; 3,915,717; 3,977,884; 3,993,801 and 3,993,491.
As is commonly known to those skilled in the electroless plating industry, commercially satisfactory electroless copper baths have required formaldehyde-type reducing agents and operate at high pH levels (11-13), using complexing agents to maintain the copper in solution. Such baths are effective from the standpoint of adequate rate of deposit, as well as quality of deposit and adherence to a substrate. Still, the baths are inherently unstable over long periods of use and require incorporation of "catalytic poisons" in carefully controlled trace amount to avoid spontaneous (bulk) decomposition. The plater must therefore always operate in a relatively narrow range between conditions which are conducive to satisfactory deposition on controlled areas of a substrate on the one hand, and random, unwanted, copper plate-out on tank walls, racks, etc., on the other. Continuous filtering of the solution and frequent cleaning of the plating tank, etc. is usually required. This is expensive in terms of time and labor, as well as in chemical component losses. Formaldehyde-type electroless copper baths are also prone to the Cannizzaro reaction, with accompanying wasted consumption of bath ingredients on that account. Additionally, formaldehyde is a volatile chemical. The bath vapors can be toxic and must accordingly be appropriately handled, which introduces environmental control problems.